Medical image segmentation - a comparison of two algorithms
Liana Stănescu, Dumitru Dan Burdescu · 2010
Image segmentation plays an important role in image analysis as a frequent pre-processing step in many image understanding algorithms and practical vision systems. According to several authors, segmentation terminates when the observer's goal is satisfied and for this reason, a unique method that can be applied to all possible cases does not yet exist. The purpose of this paper is to find which segmentation method is more appropriate for recognition and diagnosis of medical images. The algorithms used for comparison are: the color set back-projection algorithm that can be found in many related studies, and an original segmentation method using a hexagonal structure defined on the set of image pixels. Error measuring algorithms, which quantify the consistency between these two segmentations, were used in order to evaluate these segmentation methods. These measures allow a principled comparison between segmentation results on different images, with differing numbers of regions and which is generated by different algorithms with different parameters.